Lack of meta-analytic evidence for an impact of COMT Val158Met genotype on brain activation during working memory tasks.

Imaging genetics, that is, the investigation of neuroimaging correlates for genetic variation, have been highly popular over the last decade. Within this framework, schizophrenia risk genes and, in particular, the effect of the Val158Met (rs4680) single nucleotide polymorphism in the catechol-O-methyltransferase (COMT) gene on activations related to working memory have been studied extensively. COMT mediates the degradation of dopamine in the synaptic cleft, and its Val allele is associated with lower working memory performance (1). It is assumed that these effects are due to increased enzymatic activity in Val allele carriers leading to decreased prefrontal dopamine concentration, causing changes in regional activity during working memory tasks. However, is there reliable evidence to support this conclusion? To investigate whether there is converging evidence from imaging genetics studies for COMT genotype effects on working memory-related activation, we performed a quantitative coordinate-based meta-analysis using the activation likelihood estimation (ALE) approach (2). Relevant studies were retrieved through PubMed and Google Scholar, review articles, and reference tracing. Inclusion criteria are reported in the legend to Table 1. In total, 14 studies published between 2004 and 2014 met inclusion criteria (Table 1). Together, these studies enrolled a total of 995 subjects (920 healthy subjects and 75 schizophrenia patients). Although most studies tested for linear effects of increasing or decreasing Val alleles, not all publications used this contrast (cf. Table 1). To establish a standardized classification for assessing the convergence of results despite methodical heterogeneity of the original studies, we summarized different contrasts by focusing on the number of Val alleles. In the following, we thus refer to hyperor hypoactivation associated with a higher number of Val alleles in the original contrast (e.g., Met/Met . Val/Met . Val/ Val, but also, e.g., Met allele carriers . Val/Val). Among the included studies, eight studies reported only hyperactivations, whereas three studies reported only hypoactivation, and four studies reported both hyperand hypoactivation with higher numbers of Val alleles. To determine whether genotype effects showed convergent results regardless of diagnosis, both the healthy controls and the schizophrenia patients were included in a first analysis. We also conducted a subanalysis that included only the healthy controls. ALE meta-analysis across all reported findings was conducted (3), and ensuing statistical parametric maps were thresholded at p , .05 [cluster-level family-wise error (FWE), corrected for multiple comparisons, cluster-forming threshold at voxel level p , .001 (2)]. We first tested whether there were convergent results (aberrant activations) across hyperor hypoactivation associated with a higher number of Val alleles in the original contrast. We conducted additional analyses in all contrasts testing for convergence only among those